Measuring latent individual difference variables with a conjoint design and structural equation modeling

Researchers often measure consumers’ responsiveness to a certain attribute (e.g., how much consumers value sustainability in products or low prices). Traditionally, this has been done by using multi-item self-report scales. We propose the use of a rating-based conjoint design analyzed with structural equation modeling as an alternative. To illustrate the approach, we apply it to the measurement of eco-utility. In doing so, we evaluate three likely advantages of the conjoint-based approach relative to traditional self-reports: (a) reduced bias due to Acquiescence Response Style, (b) reduced bias due to socially desirable responding, and (c) enhanced ability to control for the effects of product attributes that may be confounded with the focal attribute (e.g., price may be confounded with sustainability). Our results provide preliminary support for each of the proposed advantages. We conclude with implications for measurement and discuss how the current approach offers new ideas on how to measure latent variables.